What worked for us: Accessing the the master by DNS and putting etcd on a persistent volume. This way we're able to replace the master within a DNS records ttl. As the api server is not a hard requirement for the workload, this is HA enough for us.
Can you give an example? I know next to nothing about Mesos, Aurora, or Marathon.
One area where we've been seeing some interesting work happening is in stateful services running on top of schedulers. With something like postgres, there is interesting logic that must happen for responsible management beyond just throwing up 3 of them. You need to manage replication carefully. When a node goes down somewhere, you need to react somewhere else in a controlled way. With Mesos, you can write a scheduler that encodes your runbook. With kubernetes, you need a lot more out-of-band stuff.
You're correct that the scheduler in the box is declarative. However, it is easy (well, as easy as writing a framework) to swap out a scheduler that might be a better fit for what you're looking for.
See this stackoverflow answer by one of the core Kubernetes team for more info: http://stackoverflow.com/questions/28857993/how-does-kuberne...
Mesos is much more mature, but is more low-level, and schedulers like Aurora or Marathon do not give you the same service abstraction, but they have many features not currently present in Kubernetes. But, in return for being more low-level, you can run things that are simply impossible in a declarative system like Kubernetes. Personally I see Kubernetes-on-Mesos as being a really bright future choice, but like pretty much anything other than Marathon/Aurora on Mesos, it's pretty nascent.